Metformin is the first-line agent in the treatment of type 2 diabetes, yet little is known about the molecular mechanisms of metformin action. The genetic parsing of metformin responders versus non-responders remains similarly underexplored. The Diabetes Prevention Program (DPP) includes 988 participants randomized to metformin. We propose to complete a genome-wide association study (including exome content) for metformin response in the DPP. In parallel, we will explore genetic determinants of metformin response for diabetes prevention and for lowering glycated hemoglobin: the latter will be meta-analyzed with data from 3,500 participants in the GoDARTS cohort. To prioritize human findings we will leverage information from an siRNA genomic screen for metformin response in the roundworm nematode C. elegans, integrating genes and pathways identified in the course of this screen with the human findings. Results that emerge from this integration will undergo further replication in three independent human cohorts. Top signals will be carried forward for functional validation in an established human hepatocyte model, where knockout and point mutation approaches using CRISPR technology will be used to confirm the causal gene and variant. If successful, this project will accomplish the twin goals of elucidating the mechanism of action of metformin and identifying genetic predictors of clinical response, while illustrating the arc of progress from genetic association to identifying the causal gene/variant in order to illuminate function.